Algorithmic improvements for discovery of germline copy number variants in next-generation sequencing data.

BACKGROUND: Copy number variants (CNVs) play a significant role in human heredity and disease. However, sensitive and specific characterization of germline CNVs from NGS data has remained challenging, particularly for hybridization-capture data in which read counts are the primary source of copy number information. RESULTS: We describe two algorithmic adaptations that improve CNV detection accuracy in a Hidden Markov Model (HMM) context. First, we present a method for computing target- and copy number-specific emission distributions. Second, we demonstrate that the Pointwise Maximum a posteriori (PMAP) HMM decoding procedure yields improved sensitivity for small CNV calls compared to the more common Viterbi HMM decoder. We develop a prototype implementation, called Cobalt, and compare it to other CNV detection tools using sets of simulated and previously detected CNVs with sizes spanning a single exon to a full chromosome. CONCLUSIONS: In both the simulation and previously detected CNV studies Cobalt shows similar sensitivity but significantly fewer false positive detections compared to other callers. Overall sensitivity is 80-90% for deletion CNVs spanning 1-4 targets and 90-100% for larger deletion events, while sensitivity is somewhat lower for small duplication CNVs.

Clinical utility of a next generation sequencing panel assay for Marfan and Marfan-like syndromes featuring aortopathy.

Aortopathy can be defined as aortic dilation, aneurysm, dissection, and tortuosity. Familial aortopathy may occur secondary to fibrillin-1 (FBN1) mutations in the setting of Marfan syndrome, or may occur as a result of other genetic defects with different, but occasionally overlapping, phenotypes. Because of the phenotypic overlap and genetic heterogeneity of disorders featuring aortopathy, we developed a next generation sequencing (NGS) assay and comparative genomic hybridization (CGH) array to detect mutations in 10 genes that cause thoracic aortic aneurysms (TAAs). Here, we report on the clinical and molecular findings in 175 individuals submitted for aortopathy panel testing at ARUP laboratories. Ten genes associated with heritable aortopathies were targeted using hybridization capture prior to sequencing. NGS results were analyzed, and variants were confirmed using Sanger sequencing. Array CGH was used to detect copy-number variation. Of 175 individuals, 18 had a pathogenic mutation and 32 had a variant of uncertain significance (VUS). Most pathogenic mutations (72%) were identified in FBN1. A novel large SMAD3 duplication and FBN1 deletion were identified. Over half who had TAAs or other aortic involvement tested negative for a mutation, suggesting that additional aortopathy genes exist. We anticipate that the clinical sensitivity of at least 10.3% will rise with VUS reclassification and as additional genes are identified and included in the panel. The aortopathy NGS panel aids in the timely molecular diagnosis of individuals with disorders featuring aortopathy and guides proper treatment.

Candidate locus analysis for PHACE syndrome.

PHACE syndrome (OMIM #606519) is a neurocutaneous syndrome of unknown etiology and pathogenesis. We report on an individual with PHACE syndrome with a complete deletion of SLC35B4 on 7q33. In order to further analyze this region, SLC35B4 was sequenced for 33 individuals with PHACE syndrome and one parental set. Common polymorphisms with a possible haplotype but no disease causing mutation were identified. Sixteen of 33 samples of the PHACE syndrome patients were also analyzed for copy number variations using high-resolution oligo-comparative genomic hybridization (CGH) microarray. A second individual in this cohort had a 26.5 kb deletion approximately 80 kb upstream of SLC35B4 with partial deletion of the AKR1B1 on 7q33. The deletions observed on 7q33 are not likely the singular cause of PHACE syndrome; however, it is possible that this region provides a genetic susceptibility to phenotypic expression with other confounding genetic or environmental factors.

Characterization of large genomic deletions in the FBN1 gene using multiplex ligation-dependent probe amplification.

BACKGROUND: Connective tissue diseases characterized by aortic aneurysm, such as Marfan syndrome, Loeys-Dietz syndrome and Ehlers Danlos syndrome type IV are heterogeneous and despite overlapping phenotypes, the natural history, clinical manifestations and interventional course for each diagnosis can be quite unique. The majority of mutations involved in the etiology of these disorders are missense and nonsense mutations. However, large deletions and duplications undetected by sequencing may be implicated in their pathogenesis, and may explain the apparent lack of genotype-phenotype correlation in a subset of patients. The objective of this study was to search for large pathogenic deletions and/or duplications in the FBN1, TGFßR1, and TGFßR2 genes using multiplex-ligation dependent probe amplification (MLPA) in patients with aortopathy, in whom no mutations in the FBN1, TGFßR1, and TGFßR2 genes were identified by sequencing. METHODS: The study included 14 patients from 11 unrelated families with aortic aneurysm. Of those, six patients (including 3 first-degree relatives), fulfilled the revised Ghent criteria for Marfan syndrome, and eight had predominantly aortic aneurysm/dilatation with variable skeletal and craniofacial involvement. MLPA for FBN1, TGFßR1, and TGFßR2 was carried out in all patients. A 385 K chromosome 15 specific array was used in two patients with a deletion of the entire FBN1 in order to define its size and boundaries. RESULTS: We identified two novel large deletions in the FBN1 gene in four patients of two unrelated families who met clinical diagnostic criteria for Marfan syndrome. One patient was found to have a FBN1 deletion encompassing exons 1-5. The other three patients had a 542 Kb deletion spanning the whole FBN1 gene and five additional genes (SLC24A5, MYEF2, CTXN2, SLC12A1, DUT) in the chromosome 15. CONCLUSIONS: Our findings expand the number of large FBN1 deletions, and emphasize the importance of screening for large genomic deletions in connective tissue disorders featuring aortopathies, especially for those with classic Marfan phenotype.

High-throughput amplicon scanning of the TP53 gene in breast cancer using high-resolution fluorescent melting curve analyses and automatic mutation calling.

Identifying mutations in the TP53 gene is important for cancer prognosis, predicting response to therapy, and determining genetic risk. We have developed a high-throughput scanning assay with automatic calling to detect TP53 mutations in DNA from fresh frozen (FF) and formalin-fixed paraffin-embedded (FFPE) tissues. The coding region of the TP53 gene (exons 2-11) was PCR-amplified from breast cancer samples and scanned by high-resolution fluorescent melting curve analyses using a 384-well format in the LightCycler 480 instrument. Mutations were confirmed by direct sequencing. Sensitivity and specificity of scanning and automatic mutation calling was determined for FF tissue (whole genome amplified [WGA] and non-WGA) and FFPE tissue. Thresholds for automatic mutation calling were established for each preparation type. Overall, we confirmed 27 TP53 mutations in 68 primary breast cancers analyzed by high-resolution melting curve scanning and direct sequencing. Using scanning and automatic calling, there was high specificity (>95%) across all DNA preparation methods. Sensitivities ranged from 100% in non-WGA DNA from fresh tissue to 86% in WGA DNA and DNA from formalin-fixed, paraffin-embedded tissue. Scanning could detect mutated DNA at a dilution of 1:200 in a background of wild-type DNA. Mutation scanning by high-resolution fluorescent melting curve analyses can be done in a high-throughput and automated fashion. The TP53 scanning assay can be performed from a variety of specimen types with high sensitivity/specificity and could be used for clinical and research purposes.

Expanding the clinical and molecular findings in RASA1 capillary malformation-arteriovenous malformation.

RASA1-related disorders are vascular malformation syndromes characterized by hereditary capillary malformations (CM) with or without arteriovenous malformations (AVM), arteriovenous fistulas (AVF), or Parkes Weber syndrome. The number of cases reported is relatively small; and while the main clinical features are CMs and AVMs/AVFs, the broader phenotypic spectrum caused by variants in the RASA1 gene is still being defined. Here, we report the clinical and molecular findings in 69 unrelated cases with a RASA1 variant identified at ARUP Laboratories. Sanger sequencing and multiplex ligation-dependent probe amplification were primarily used to evaluate RASA1. Several atypical cases were evaluated using next-generation sequencing (NGS) and array-comparative genomic hybridization (aCGH). Sixty individuals had a deleterious RASA1 variant of which 29 were novel. Nine individuals had a variant of uncertain significance. Five large RASA1 deletions were detected, giving an overall deletion/duplication rate of 8.3% (5/60) among positive cases. Most (75.4%) individuals with a RASA1 variant had CMs, and 44.9% had an AVM/AVF. Clinical findings in several cases expand the RASA1 phenotype. Our data suggest that screening for large RASA1 deletions and duplications in this disorder is important and suggest that NGS multi-gene panel testing is beneficial for the molecular diagnosis of cases with complex vascular phenotypes.

A Copy Number Variant on Chromosome 20q13.3 Implicated in Thinness and Severe Obesity.

BACKGROUND/OBJECTIVES: To identify copy number variants (CNVs) which are associated with body mass index (BMI). SUBJECTS/METHODS: CNVs were identified using array comparative genomic hybridization (aCGH) on members of pedigrees ascertained through severely obese (BMI ≥ 35 kg/m(2)) sib pairs (86 pedigrees) and thin (BMI ≤ 23 kg/m(2)) probands (3 pedigrees). Association was inferred through pleiotropy of BMI with CNV log⁡2 intensity ratio. RESULTS: A 77-kilobase CNV on chromosome 20q13.3, confirmed by real-time qPCR, exhibited deletions in the obese subjects and duplications in the thin subjects (P = 2.2 × 10(-6)). Further support for the presence of a deletion derived from inference by likelihood analysis of null alleles for SNPs residing in the region. CONCLUSIONS: One or more of 7 genes residing in a chromosome 20q13.3 CNV region appears to influence BMI. The strongest candidate is ARFRP1, which affects glucose metabolism in mice.

Processed Pseudogene Confounding Deletion/Duplication Assays for SMAD4.

Mutations in SMAD4 have been associated with juvenile polyposis syndrome and combined juvenile polyposis/hereditary hemorrhagic telangiectasia syndrome. SMAD4 is part of the SMAD gene family. To date, there has been no report in the literature of a SMAD4 pseudogene. An unusual SMAD4 duplication pattern was seen in multiple patient samples using two different duplication/deletion platforms: multiplex ligation-dependent probe amplification and chromosomal microarray. Follow-up confirmatory testing included real-time quantitative PCR and sequencing of an exon/exon junction, all results leading to the conclusion of the existence of a processed pseudogene. Examination of clinical results from two laboratories found a frequency of 0.26% (12 in 4672 cases) for this processed pseudogene. This is the first report of the presence of a processed pseudogene for SMAD4. We believe that knowledge of its existence is important for accurate interpretation of clinical diagnostic test results and for new assay designs. This study also indicates how a processed pseudogene may confound quantitative results, dependent on placement of probes and/or primers in a particular assay design, potentially leading to both false-positive and false-negative results. We also found that the SMAD4 processed pseudogene affects next-generation sequencing results by confounding the alignment of the sequences, resulting in erroneous variant calls. We recommend Sanger sequencing confirmation for SMAD4 variants.

Pregnancy outcome in recurrent miscarriage patients with skewed X chromosome inactivation.

OBJECTIVE: To analyze X inactivation in women with recurrent miscarriage to estimate whether skewed X inactivation is associated with recurrent miscarriage and whether it predicts next pregnancy outcomes. METHODS: A multicenter study was performed. A power calculation determined that 101 patients were needed to detect a difference in skewed X inactivation between patients and controls. Patients were entered into a prospective trial of mononuclear-cell immunotherapy and subsequently tested for skewed X inactivation. Age-matched controls had one live birth and no prior miscarriages. Results from our X inactivation assay were compared with those from an independent genetics laboratory. RESULTS: Greater than 75% skewing was seen in 22.6% of patients and 26.5% controls (P =.52). Greater than 90% skewing was seen in 6.6% of patients and 3.9% of controls (P =.77). There were 19.8% of primary aborters and 32% of secondary aborters with greater than 75% skewed X inactivation (P =.38). There were 4.9% of primary aborters and 12.0% of secondary aborters with greater than 90% skewed X inactivation (P =.27) Neither greater than 75% nor greater than 90% skewed X inactivation impacted next pregnancy outcomes (odds ratios = 0.87 [95% confidence interval (CI) 0.34, 2.3] and 1.4 [95% CI 0.27, 7.5], respectively). Results of the exchange of samples with an independent laboratory were highly correlated (alpha = 0.987, P <.001, coefficient of variation = 5.5%). CONCLUSION: Skewed X chromosome inactivation is not associated with recurrent miscarriage. A patient's X chromosome inactivation status does not predict next pregnancy outcome. Our assay correlates with another experienced laboratory.

Noncontinuously binding loop-out primers for avoiding problematic DNA sequences in PCR and sanger sequencing.

We present a method in which noncontinuously binding (loop-out) primers are used to exclude regions of DNA that typically interfere with PCR amplification and/or analysis by Sanger sequencing. Several scenarios were tested using this design principle, including M13-tagged PCR primers, non-M13-tagged PCR primers, and sequencing primers. With this technique, a single oligonucleotide is designed in two segments that flank, but do not include, a short region of problematic DNA sequence. During PCR amplification or sequencing, the problematic region is looped-out from the primer binding site, where it does not interfere with the reaction. Using this method, we successfully excluded regions of up to 46 nucleotides. Loop-out primers were longer than traditional primers (27 to 40 nucleotides) and had higher melting temperatures. This method allows the use of a standardized PCR protocol throughout an assay, keeps the number of PCRs to a minimum, reduces the chance for laboratory error, and, above all, does not interrupt the clinical laboratory workflow.

Verification of multiplex ligation-dependent probe amplification probes in the absence of positive samples.

Deletions and duplications of single or multiple exons in specific genes are associated with human diseases. Multiplex ligation-dependant probe amplification (MLPA), a technique recently introduced to clinical laboratories, can detect deletions or duplications at the exon level. MLPA kits have a high multiplexing capability containing mixtures of exon-specific probes that target the gene of interest and control probes that hybridize to other genomic areas before PCR amplification. To verify each probe set, known positive samples with a single-exon deletion or duplication and normal samples are ideally used. Often, positive samples do not exist for each exon and normal samples are not suited to verify the identity of each probe set. We designed a straightforward approach using mixes of exon-specific PCR products as template to unequivocally verify each probe set in MLPA kits. This method can be used to verify the identity of MLPA probes for exons when positive samples are unavailable. Exon-specific probes from 15 MLPA kits were shown to hybridize to the targeted exons of interest. In one kit, this method identified probes that also bind a pseudogene, making them unreliable for clinical analysis. Incorporating this methodology in the analytical validation process will help ensure that MLPA results are interpreted correctly.

Differentiating Ewing's sarcoma from other round blue cell tumors using a RT-PCR translocation panel on formalin-fixed paraffin-embedded tissues.

Ewing's sarcoma is a common malignancy of bone and soft tissue that occurs most often in children and young adults. Differentiating Ewing's sarcoma from other round blue cell tumors can be a diagnostic challenge because of their similarity in histology and clinical presentation. Thus, ancillary molecular tests for detecting disease-defining translocations are important for confirming the diagnosis. We analyzed 65 round blue cell tumors, including 53 Ewing's sarcoma samples from 50 unique cases. Samples were processed for RNA from archived formalin-fixed paraffin-embedded tissue blocks. Real-time RT-PCR assays specific for Ewing's sarcoma (EWS-FLI1, EWS-ERG, EWS-ETV1, EWS-ETV4, and EWS-FEV), synovial sarcoma (SYT-SSX1 and SYT-SSX2), and rhabdomyosarcoma (PAX3-FKHR and PAX7-FKHR) were tested across the samples. The translocation panel had a sensitivity of 81% (43 out of 53 samples) for diagnosing Ewing's sarcoma when using the histological criteria as the 'gold' standard. None of the Ewing's specific translocations were found in the non-Ewing's samples (100% specificity). Of the 43 samples with translocations detected, 26 (60%) had an EWS-FLI1 type 1 translocation, 13 (30%) had an EWS-FLI1 type 2 translocation, 3 (7%) had an EWS-ERG translocation, 1 had an EWS-ETV1 translocation, and 1 sample had both an EWS-FLI1 type 1 and type 2 translocation. Our real-time RT-PCR assay for detecting sarcoma translocations has high sensitivity and specificity for Ewing's sarcoma and has clinical utility in differentiating small round blue cell tumors in the clinical lab.

Molecular classification of melanoma using real-time quantitative reverse transcriptase-polymerase chain reaction.

BACKGROUND: The early detection and characterization of metastatic melanoma are important for prognosis and management of the disease. Molecular methods are more sensitive in detecting occult lymph node metastases compared with standard histopathology and are reported to have utility in clinical diagnostics. METHODS: Using real-time quantitative reverse transcriptase-polymerase chain reaction ([q]RT-PCR), the authors examined 36 samples (30 melanomas, 4 benign nevi, and 2 reactive lymph nodes) for the expression of 20 melanoma-related genes that function in cell growth and differentiation (epidermal growth factor receptor [EGFR], WNT5A, BRAF, FOS, JUN, MATP, and TMP1), cell proliferation (KI-67, TOP2A, BUB1, BIRC5, and STK6), melanoma progression (CD63, MAGEA3, and GALGT), and melanin synthesis (TYR, MLANA, SILV, PAX3, and MITF). In addition, samples were tested for mutations in BRAF (exons 11 and 15) and NRAS (exons 2 and 3). RESULTS: Hierarchical clustering analysis of the expression data was able to distinguish between the melanoma and nonmelanoma samples and further stratified the melanoma samples into two groups differentiated by high expression of the genes involved in beta-catenin activation (EGFR and WNT5A) and the MAPK/ERK pathway (BRAF, FOS, and JUN). Eighteen of the 28 patients (64%) were found to have mutations in either exon 15 of BRAF (V599 substitution) or codon 61 of NRAS. The mutations were mutually exclusive and did not appear to be associated with the different expression subtypes. CONCLUSIONS: The results of the current study demonstrate that real-time qRT-PCR can be analyzed using hierarchical clustering to identify expression patterns that differentiate between melanomas and other tissue types. Using a supervised analysis of the data, the authors found that the best discriminators for molecularly distinguishing between melanoma, benign nevi, and lymph nodes were MLANA, CD63, and BUB1. These markers could have diagnostic utility for the detection of melanoma micrometastasis in sentinel lymph nodes.